What is the endocannabinoid system?

This is the part of the biology that explains why a plant compound has any effect on human beings at all. It is also the part most heavily over-interpreted in marketing, so it is worth separating what is established from what is speculation.

What it is

A signalling system made up of three parts: molecules the body produces, receptors those molecules act on, and enzymes that build and break them down. It was discovered because researchers were trying to work out how THC acts.

The two best-described signalling molecules are anandamide and 2-AG. Unlike most neurotransmitters, they are made on demand rather than stored, and they act backwards across a synapse — the receiving cell releases them to tell the sending cell to ease off.

That backwards, on-demand design is the clue to what the system does. It is a regulator. It does not run any single process; it adjusts the intensity of other processes — pain signalling, appetite, mood, sleep, immune activity, memory. Think of it less as an engine and more as a set of dimmer switches, nudging other systems up or down to keep them in balance. That role as a fine-tuner, rather than a driver, is worth holding on to, because it is the part most often lost when the system is described in marketing.

The two receptors

CB1 receptors are found in high density in the brain and central nervous system. CB2 receptors are found mainly in immune cells and peripheral tissue.

• CB1 — abundant in brain regions handling memory, movement, appetite and pain processing. THC's action here explains both its therapeutic effects and its intoxicating and cognitive ones.
• CB2 — concentrated in immune tissue and peripheral cells. Interest here centres on inflammation, and drugs acting selectively at CB2 have been an active research target precisely because they would not cause intoxication. That research is promising in principle but has yet to translate into widely licensed medicines, which is a useful illustration of how far a plausible target can be from a proven treatment.

Notably, CB1 receptors are sparse in the brainstem regions that control breathing. That is the accepted explanation for why cannabinoid overdose does not cause the respiratory depression that makes opioid overdose lethal — a real and important difference, and one of the few places where a favourable comparison is warranted. It should not be read as a claim that cannabis-based medicines are harmless. Their risks are real, but they are of a different kind: impairment, anxiety, effects on the heart and, over time, tolerance and dependence, rather than the sudden lethal effect on breathing seen with opioids.

Where the claims outrun the evidence

Two ideas circulate widely and neither is established. It is worth recognising them, because they are often used to argue that almost any condition might respond.

The first is clinical endocannabinoid deficiency — the proposal that conditions such as migraine, fibromyalgia and IBS share an underlying shortfall in endocannabinoid signalling. It is a legitimate research hypothesis that has been discussed for over two decades. It is not a diagnosis, there is no test for it, and no treatment is licensed on the basis of it. The distinction matters because the phrase is sometimes used as though it were an established condition that cannabis then corrects, which would imply a much stronger case for treatment than the evidence supports. Being told you have a deficiency, and being told a particular product will fix it, are both steps well beyond what is known.

The second is the argument from ubiquity: because the system touches many functions, cannabis must help many conditions. That does not follow. Paracetamol acts on widely distributed systems too, and is useful for a narrow range of things. Breadth of receptor distribution predicts breadth of side effects at least as reliably as it predicts benefit. In fact the same reach that makes the system sound like a cure-all is exactly why cannabis-based medicines cause the range of effects they do, wanted and unwanted alike — a system involved in memory, movement, mood and appetite cannot be nudged in one place without touching the others.

What the evidence supports condition by condition is set out in Can medical cannabis treat my condition?

How the system relates to treatment in practice

Understanding the endocannabinoid system explains why cannabis-based medicines have any effect at all, but it does not, on its own, predict whether they will help a given person with a given condition. Mechanism and demonstrated benefit are separate questions.

THC produces its effects by mimicking the body's own signalling molecules at the CB1 receptor. That is a real and well-characterised mechanism, and it is the reason a plant compound can alter pain, appetite, sleep and mood. But a plausible mechanism is only a starting point. Medicine is full of interventions that made good biological sense and still did not work in patients, which is why regulators and specialists look for evidence of outcomes rather than mechanism alone.

In practice this means treatment is approached cautiously and individually. Where a specialist judges a cannabis-based product appropriate, it is usually started at a low dose and increased slowly under supervision, with follow-up reviews to check whether it is helping and whether side effects are tolerable. The medicines are typically oils or capsules taken by mouth, or dried flower for vaporising. How the dose is arrived at and adjusted is set out in How is a dose decided and adjusted?, and the difference between the two main compounds is covered in What is the difference between THC and CBD?

Reading endocannabinoid claims critically

Because the system touches so many functions, it is often invoked to suggest cannabis can treat almost anything. A few questions help separate established biology from marketing.

• Biology or outcome? Is the claim about how the system works, or about a treatment result in patients? The first can be well established while the second is not.
• Deficiency as diagnosis? If "endocannabinoid deficiency" is being presented as a condition, remember it is a research hypothesis, with no test and no licensed treatment.
• Argument from ubiquity? A claim that rests on the system being everywhere in the body proves little — wide receptor distribution predicts side effects at least as reliably as benefit.
• Human evidence? Is there controlled evidence in people for the specific condition, or only laboratory or animal work?

What matters for your own case is not the elegance of the mechanism but an individual assessment by a specialist who has seen your records. Whether treatment is appropriate is decided condition by condition and person by person, not from the biology in the abstract, and some people are advised that it is not suitable for them.

Patient guide · Written by the Flora House editorial team · Last reviewed: 7 October 2026